• The Journal of the Acoustical Society of Korea
• /
• v.41 no.6
• /
• pp.629-636
• /
• 2022

We present a preprocessing method for an underwater target detection model based on a convolutional neural network. The acoustic characteristics of the ship show ambiguous expression due to the strong signal power of the low frequency. To solve this problem, we combine feature preprocessing methods with various feature scaling methods and spectrogram methods. Define a simple convolutional neural network model and train it to measure preprocessing performance. Through experiment, we found that the combination of log Mel-spectrogram and standardization and robust scaling methods gave the best classification performance.

• International journal of advanced smart convergence
• /
• v.12 no.4
• /
• pp.378-385
• /
• 2023

The importance of wearing respiratory protective equipment has been highlighted even more during the COVID-19 pandemic. Even if the suitability of respiratory protection has been confirmed through testing in a laboratory environment, there remains the potential for leakage points in the respirators due to improper application by the wearer, damage to the equipment, or sudden movements in real working conditions. In this paper, we propose a method to detect the occurrence of leak holes by measuring the pressure changes inside the mask according to the wearer's breathing activity by attaching an IoT sensor to a full-face respirator. We designed 9 experimental scenarios by adjusting the degree of leak holes of the respirator and the breathing cycle time, and acquired respiratory data for the wearer of the respirator accordingly. Additionally, we analyzed the respiratory data to identify the duration and pressure change range for each breath, utilizing this data to train a neural network model for detecting leak holes in the respirator. The experimental results applying the developed neural network model showed a sensitivity of 100%, specificity of 94.29%, and accuracy of 97.53%. We conclude that the effective detection of leak holes can be achieved by incorporating affordable, small-sized IoT sensors into respiratory protective equipment.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.26 no.5
• /
• pp.422-430
• /
• 2023

To build an automated system detecting toxic chemicals from Raman spectra, we have to obtain sufficient data of toxic chemicals. However, it usually costs high to gather Raman spectra of toxic chemicals in diverse situations. Tackling this problem, we develop methods to generate synthetic Raman spectra of DMMP and 2-CEES without actual experiments. First, we propose certain mathematical transforms to augment few original Raman spectra. Then, we train deep generative models to generate more realistic and diverse data. Analyzing synthetic Raman spectra of toxic chemicals generated by our methods through visualization, we qualitatively verify that the data are sufficiently similar to original data and diverse. For conclusion, we obtain a synthetic dataset of DMMP and 2-CEES with the proposed algorithm.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.1-23
• /
• 2018

From the 21st century, various high-quality services have come up with the growth of the internet or 'Information and Communication Technologies'. Especially, the scale of E-commerce industry in which Amazon and E-bay are standing out is exploding in a large way. As E-commerce grows, Customers could get what they want to buy easily while comparing various products because more products have been registered at online shopping malls. However, a problem has arisen with the growth of E-commerce. As too many products have been registered, it has become difficult for customers to search what they really need in the flood of products. When customers search for desired products with a generalized keyword, too many products have come out as a result. On the contrary, few products have been searched if customers type in details of products because concrete product-attributes have been registered rarely. In this situation, recognizing texts in images automatically with a machine can be a solution. Because bulk of product details are written in catalogs as image format, most of product information are not searched with text inputs in the current text-based searching system. It means if information in images can be converted to text format, customers can search products with product-details, which make them shop more conveniently. There are various existing OCR(Optical Character Recognition) programs which can recognize texts in images. But existing OCR programs are hard to be applied to catalog because they have problems in recognizing texts in certain circumstances, like texts are not big enough or fonts are not consistent. Therefore, this research suggests the way to recognize keywords in catalog with the Deep Learning algorithm which is state of the art in image-recognition area from 2010s. Single Shot Multibox Detector(SSD), which is a credited model for object-detection performance, can be used with structures re-designed to take into account the difference of text from object. But there is an issue that SSD model needs a lot of labeled-train data to be trained, because of the characteristic of deep learning algorithms, that it should be trained by supervised-learning. To collect data, we can try labelling location and classification information to texts in catalog manually. But if data are collected manually, many problems would come up. Some keywords would be missed because human can make mistakes while labelling train data. And it becomes too time-consuming to collect train data considering the scale of data needed or costly if a lot of workers are hired to shorten the time. Furthermore, if some specific keywords are needed to be trained, searching images that have the words would be difficult, as well. To solve the data issue, this research developed a program which create train data automatically. This program can make images which have various keywords and pictures like catalog and save location-information of keywords at the same time. With this program, not only data can be collected efficiently, but also the performance of SSD model becomes better. The SSD model recorded 81.99% of recognition rate with 20,000 data created by the program. Moreover, this research had an efficiency test of SSD model according to data differences to analyze what feature of data exert influence upon the performance of recognizing texts in images. As a result, it is figured out that the number of labeled keywords, the addition of overlapped keyword label, the existence of keywords that is not labeled, the spaces among keywords and the differences of background images are related to the performance of SSD model. This test can lead performance improvement of SSD model or other text-recognizing machine based on deep learning algorithm with high-quality data. SSD model which is re-designed to recognize texts in images and the program developed for creating train data are expected to contribute to improvement of searching system in E-commerce. Suppliers can put less time to register keywords for products and customers can search products with product-details which is written on the catalog.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.10
• /
• pp.2461-2469
• /
• 1997

Target detection and recognition problems, in which neural networks are widely used, require translation invariant and real-time processing in addition to the requirements that general pattern recognition problems need. This paper presents a novel architecture that meets the requirements and explains effective methodology to train the network. The proposed neural network is an architectural extension of the shared-weight neural network that is composed of the feature extraction stage followed by the pattern recognition stage. Its feature extraction stage performs correlational operation on the input with a weight kernel, and the entire neural network can be considered a nonlinear correlation filter. Therefore, the output of the proposed neural network is correlational plane with peak values at the location of the target. The architecture of this neural network is suitable for implementing with parallel or distributed computers, and this fact allows the application to the problems which require realtime processing. Net training methodology to overcome the problem caused by unbalance of the number of targets and non-targets is also introduced. To verify the performance, the proposed network is applied to detection and recognition problem of a specific automobile driving around in a parking lot. The results show no false alarms and fast processing enough to track a target that moves as fast as about 190 km per hour.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.1
• /
• pp.55-64
• /
• 2013

Due to the popularization of digital high-performance capturing equipments and the emergence of powerful image-editing softwares, it is easy for anyone to make a high-quality counterfeit money. However, the probability of detecting a counterfeit money to the general public is extremely low. In this paper, we propose a counterfeit money detection algorithm using a general purpose scanner. This algorithm determines counterfeit money based on the different features in the printing process. After the non-local mean value is used to analyze the noises from each money, we extract statistical features from these noises by calculating a gray level co-occurrence matrix. Then, these features are applied to train and test the support vector machine classifier for identifying either original or counterfeit money. In the experiment, we use total 324 images of original money and counterfeit money. Also, we compare with noise features from previous researches using wiener filter and discrete wavelet transform. The accuracy of the algorithm for identifying counterfeit money was over 94%. Also, the accuracy for identifying the printing source was over 93%. The presented algorithm performs better than previous researches.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.2
• /
• pp.139-147
• /
• 2021

This paper introduces and implements a Sound Event Detection (SED) system based on weakly-supervised learning where only part of the data is labeled, and analyzes the effect of parameters. The SED system estimates the classes and onset/offset times of events in the acoustic signal. In order to train the model, all information on the event class and onset/offset times must be provided. Unfortunately, the onset/offset times are hard to be labeled exactly. Therefore, in the weakly-supervised task, the SED model is trained by "strongly labeled data" including the event class and activations, "weakly labeled data" including the event class, and "unlabeled data" without any label. Recently, the SED systems using the mean-teacher model are widely used for the task with several parameters. These parameters should be chosen carefully because they may affect the performance. In this paper, performance analysis was performed on parameters, such as the feature, moving average parameter, weight of the consistency cost function, ramp-up length, and maximum learning rate, using the data of DCASE 2020 Task 4. Effects and the optimal values of the parameters were discussed.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.4
• /
• pp.329-344
• /
• 2023

In this study, we comprehensively analyze the generalization performance of various deep learning-based active sonar target classifiers when applied to small and imbalanced active sonar datasets. To generate the active sonar datasets, we use data from two different oceanic experiments conducted at different times and ocean. Each sample in the active sonar datasets is a time-frequency domain image, which is extracted from audio signal of contact after the detection process. For the comprehensive analysis, we utilize 22 Convolutional Neural Networks (CNN) models. Two datasets are used as train/validation datasets and test datasets, alternatively. To calculate the variance in the output of the target classifiers, the train/validation/test datasets are repeated 10 times. Hyperparameters for training are optimized using Bayesian optimization. The results demonstrate that shallow CNN models show superior robustness and generalization performance compared to most of deep CNN models. The results from this paper can serve as a valuable reference for future research directions in deep learning-based active sonar target classification.

• Journal of Preventive Medicine and Public Health
• /
• v.20 no.2 s.22
• /
• pp.312-321
• /
• 1987

To investigate the maternal and child factors associated with early detection of cerebral palsy, 74 mothers of cerebral palsy children who were born since January 1, 1980 and being treated at Taegu Rehabilitation Center for the Handicapped, Rehabilitiation Center of Taegu University, St. Paul Children's House and Pusan Welfare Association of Cerebral Palsy Children were interviewed from February to April 1987. There is no association between age of child when parents noticed the child's abnormality and educational level of father but it tend to be detected earlier when education level of mother is college or above compared with high school or under. There is a trend of earlier detection of child's abnormality although statistically not significant in case father is professional or managerial worker, monthly income of father is over \610,000, child is first-born, age of the parents is 34 years or under, child is a boy, and child has periodic well-baby check-up. The child's abnormality is detected earlier when mothers had 7 prenatal visits or more compared with those who had 6 visits or less (p<0.05). Parents noticed the child's abnormality first in 85.1% of the cases whereas doctors detected it first in 2.7% and this percentage was not different whether the child had periodic well-baby check-up or not. The first physician's diagnosis of the children was cerebral palsy in 36.5% and the rest was normal, need for observation, uncertain, etc. Parents took the child to doctor for diagnosis 2-3 months after they noticed the child's abnormality and after the child was diagnosed as cerebral palsy parents either took no therapeutic measure or brought the child to physiotherapy or acupuncture or gave herb medicine before they started specific rehabilitative therapy. For early detection of the cerebral palsy children, teaching of evaluation method for child development should be reinforced both in medical school and clinical training course and should train the specialist for diagnosis and treatment of crippling conditions. Also, public education is needed for the importance of early detection of crippling conditions and currently available methods for diagnosis and treatment.

• Tunnel and Underground Space
• /
• v.31 no.5
• /
• pp.374-384
• /
• 2021

In this study, we propose a new approach for automatic fracture detection in CT scan images of rock specimens. This approach is built on top of two-stage object detection deep learning algorithm called Faster R-CNN with a major modification of using rotated bounding box. The use of rotated bounding box plays a key role in the future work to overcome several inherent difficulties of fracture segmentation relating to the heterogeneity of uninterested background (i.e., minerals) and the variation in size and shape of fracture. Comparing to the commonly used bounding box (i.e., axis-align bounding box), rotated bounding box shows a greater adaptability to fit with the elongated shape of fracture, such that minimizing the ratio of background within the bounding box. Besides, an additional benefit of rotated bounding box is that it can provide relative information on the orientation and length of fracture without the further segmentation and measurement step. To validate the applicability of the proposed approach, we train and test our approach with a number of CT image sets of fractured granite specimens with highly heterogeneous background and other rocks such as sandstone and shale. The result demonstrates that our approach can lead to the encouraging results on fracture detection with the mean average precision (mAP) up to 0.89 and also outperform the conventional approach in terms of background-to-object ratio within the bounding box.